Effective Data Handling

Like many IT professionals of my generation, I began programming with COBOL on mainframes. One thing that I always found odd about COBOL was its incredibly relaxed rules on datatypes (especially after having the importance of structured data typing beaten into me through the long years of university education). When processing a file with COBOL, a sequence of bytes read from a file could be treated as a character string, a date, a number, even binary data—as long as I could overlay the bytes successfully into a working storage definition then the COBOL compiler was happy.